|Abstract:||Angiogenesis is a key regulator in over 70 diseases, and it is typically characterized via uni-family ligand-receptor interactions—primarily via VEGF-A:VEGFR2 binding. But targeting the VEGF family alone has not achieved the promise of stable vascular control, because angiogenesis involves other signaling axes, such as the PDGFs. In cancer, for example, anti- angiogenic therapeutics primarily target the VEGF signaling family, but their success is limited by the development of drug resistance. These challenges in VEGF targeting, and recent discoveries of VEGF-A:PDGFR binding and signaling, presents a compelling need to shift away from a uni-family (e.g., VEGF-alone) towards a multi- and cross-family (e.g., VEGF + PDGF, etc.) understanding of angiogenesis.
VEGFs and PDGFs share significant structural and binding motifs, which could explain the new VEGF-A: PDGFR interaction and suggests that additional cross-family VEGF/PDGF cross-family binding may occur. To identify and measure new ligand:receptor binding pairs, I utilized a surface plasmon resonance-biosensor based kinetics assay. Here, I present my discovery of novel PDGF:VEGFR2 interactions, with: PDGF-AA:R2 KD = 530 nM, PDGF- AB:R2 KD = 110 pM, PDGF-BB:R2 KD = 40 nM, and PDGF-CC:R2 KD = 70 pM. For the first time, I measured the kinetic constants for VEGF-A:PDGFRβ and the canonical PDGF:PDGFR interactions. I then construct a model of VEGF-A: and PDGF:VEGFR2 interactions in endothelial cells (ECs),and predict that PDGF:VEGFR binding could contribute up to 96% of VEGFR2 ligation in healthy conditions and in cancer.
Because VEGFR2-mediated signaling drives angiogenic responses, I investigated whether PDGF-binding could activate VEGFR2-mediated signaling, and ultimately stimulate EC proliferation and migration—steps key in angiogenesis. I found that PDGFs induce significant VEGFR2 phosphorylation at tyrosines 951, 1054/59, and 1175. Further, PDGFs activate the downstream signaling effectors FAK, PI3K, PLCγ and Src. Remarkably, PDGF-AA promoted a 1.2-fold larger increase in Y1054/59 phosphorylation than did the canonical VEGF-A. PDGF- BB stimulates a ~1.3-fold larger increase in FAK phosphorylation than VEGF-A. Furthermore, I show the PDGFs stimulate increased EC proliferation, and that only PDGF-AA, -BB, and -CC stimulated significant EC migration. Taken together, these findings that PDGFs can stimulate VEGFR2 signaling and downstream angiogenic cell responses will break new ground towards modulating angiogenesis in health and disease, offering new hypotheses for why angiogenesis- targeting therapies have proven unsuccessful.
VEGF-A165 and its binding to VEGFRs is an important angiogenesis regulator. But several additional splice variants play prominent roles in regulating angiogenesis in health and in vascular disease, including VEGF-A121 and an anti-angiogenic variant VEGF-A165b. Despite their significance to physiological, therapeutic, and pathological angiogenesis, little is known about differences in their binding. I measured the binding kinetics for VEGF-A165, -A165b, and -A121 with VEGFR1 and –R2 using (SPR). I find that the VEGF-A variants differentially bind VEGFRs. VEGFR1 binds variants at strengths: VEGF-A165a > -A165b > -A121; and VEGFR2 binds variants at strengths: VEGF-A165b > -A165a > -A121. Interestingly, my findings suggest that the anti-angiogenic VEGF-A165a would preferentially bind VEGFR2, out-competing the pro- angiogenic isoform. My results suggest VEGF-A splice variants may play an important modulatory role in VEGFR-mediated angiogenesis, stressing the need to differentiate VEGF-A splice variants in future studies.
Overall, my results demonstrate that that VEGFR-mediated angiogenesis involves a more complex network of ligand:receptor interactions than previously known. Computational approaches can provide key insight to detangle these signaling pathways but are limited by the sparse knowledge of cross-family interactions. Here, I present a framework for studying known and potential non-canonical interactions. I construct generalized models of RTK ligation and dimerization for systems of two, three, and four receptor types, and different degrees of cross- family ligation. Across each model, I develop parameter-space maps that fully determine relative receptor activation for any set of ligand:receptor binding kinetics, ligand concentrations, and receptor concentrations. My generalized models serve as a powerful reference tool for predicting dimerization for known ligand:receptor axes. They also can predict how unknown interactions could alter signaling dimerization patterns. Accordingly, they will drive exploration of cross- family interactions, and help guide therapeutic developments across processes like cancer and cardiovascular diseases that depend on RTK-mediated signaling.